Meta treats a comment on a Facebook page post almost exactly the way it treats a comment on an Instagram post. DataReportal counted 7.01 million social media user identities in Israel in October 2025, 73.4 percent of the population, and a large share of those people hold both accounts. Almost nobody runs the Facebook half.
The reason is not technical. It is that every guide about this mechanic was written about Instagram, so page owners assume it is an Instagram feature that Facebook happens to host.
It is the other way round. The mechanic is a page capability, and Instagram borrows it.
What is identical and what actually differs
Identical: the permission model, the keyword match, and the pairing of a public reply with a private message. Different: where the message lands, what counts as an eligible post, and how much comment volume there is.
If you already understand the Instagram version, you know most of this. The full mechanics of keyword choice, the public reply and the private message are in how the Instagram reply for link funnel actually works, and none of it changes here.
| Piece | Facebook page | Instagram professional account |
|---|---|---|
| Permission source | Granted on the page | Granted on the linked page |
| Where the message lands | Messenger thread, ordinary notification | Inbox for followers, message requests for everyone else |
| Eligible surfaces | Page posts, page Reels, finished live videos | Feed posts, Reels |
| Excluded | Group posts, stories, shared posts from other pages | Stories, other accounts posts |
| Typical comment volume | Lower | Higher |
Read that table as a trade. Facebook gives you better delivery on fewer comments, Instagram gives you more comments with worse delivery. Neither is the better platform in the abstract, and the answer depends entirely on who your buyer is.
What a page needs before it can run this
Three things: a Facebook page you administer, a connection granted through Meta official permission flow, and Messenger enabled on the page. No password is handed over, and nothing is scraped.
The person connecting has to hold admin rights on the page, not editor rights. If the page sits inside a Business Manager, the permission has to be granted at that level, which is where most failed connections come from. The symptom is an empty page list at the moment you are asked to choose one.
Messenger has to be switched on for the page. Some pages, particularly ones set up years ago by an agency, have messaging disabled, and a private reply to a comment has nowhere to go when it is.
The permissions themselves are the ordinary set: read comments on the page posts, publish a reply, and send a message. That is what the consent screen shows you before you approve it, and it is worth reading rather than clicking through.
If you also want the Instagram side, the professional account has to be linked to that same page. That link is the whole reason one connection can cover both, and it is the first thing to check when Instagram automations do not appear after a successful Facebook connection.
How a private reply behaves on Facebook against Instagram
On Facebook, the private reply opens a Messenger thread and arrives as a normal notification. On Instagram, the same message lands in the inbox for followers and in message requests for everyone else, where much of it is never seen.
That difference is the single strongest argument for running the Facebook half. On Instagram you have to write the public reply so it tells people to go looking in their requests folder. On Facebook you do not, because Messenger simply notifies them.
Meta documentation describes the Facebook private reply as one per comment, sent within a window measured in days from the comment rather than at any time you like. Treat that as a summary rather than a quotation, and check the current developer docs before building anything that depends on the exact limit.
The practical consequence of one reply per comment is that you get a single shot per commenter, per comment. If your message is three paragraphs with the link at the bottom, that shot is wasted. One line of context, one link, one question.
The ordinary messaging window then applies as usual. Once the person answers, you have a normal conversation for the standard period, and past that you are back to approved message types. So the useful work happens in the first exchange, which is also when the person actually cares.
Groups, and why a group is not a page
A group is not a page. These permissions are granted per page and cover posts the page owns. A post inside a group, including one you run, is not a page post, so the mechanic does not reach it.
There is a reason beyond the API surface. A group belongs, socially, to its members. People post there under an expectation that a business is not harvesting the thread, and an automated private message to a group member who commented is exactly the behaviour that produces reports.
Meta has also wound down what third party tools can do with groups over several years, so anything you read about group automation from a few years ago should be assumed out of date.
The workable pattern is the boring one: publish the offer as a page post, run the automation there, and use the group to point members at it. The comment then happens on the post, where the permission actually exists, and the group stays a group.
If your audience genuinely lives in a group and not on your page, that is a signal about where to publish, not a reason to automate inside the group.
Which post types this works on
Anything on the page that carries a comment thread and belongs to the page: photo posts, video posts, link posts, plain text posts, page Reels and live videos once they have ended. That covers nearly everything a page publishes.
Boosted posts are included, because a boosted page post is still a page post and the comments on it are comments on that post. This is the cheapest way to add volume to a mechanic that is already working organically.
What is excluded is worth knowing before you plan a campaign around it. Stories carry no comment thread at all. A post you shared from another page belongs to that page, and the comments on the original are not yours to act on.
Live video is the interesting case. Comments during a live stream arrive faster than anything else a page produces, and the post remains on the page afterwards with the thread attached, so a keyword announced on air keeps working for weeks after the broadcast ends.
Why Facebook still matters for an audience that has aged with it
Because for a great many businesses the buyer is not twenty two. People who joined Facebook in their twenties are now in their forties, holding budgets, and many never moved their primary account anywhere else.
That is an argument about your audience, not a general claim about platforms. The way to settle it is to look at where your actual enquiries came from over the last six months, rather than at where the industry says attention is.
There is a second argument, and it is about competition rather than size. The comment to DM mechanic is saturated on Instagram, where audiences have seen it hundreds of times. On a Facebook page it is still unusual enough that people read the public reply.
The volume context is worth keeping in view. Socialinsider benchmark study of Instagram, built on 35 million posts across 447,613 active pages from January to December 2025, reports average engagement at 0.48 percent, down 24 percent year on year, with a median of 3 comments on a reel for accounts of 1,000 to 5,000 followers and 12 at 10,000 to 50,000.
Those are Instagram numbers, and Instagram is the busier of the two. If a median Instagram post produces a handful of comments, the Facebook half is not a rounding error you can ignore. It is a comparable handful, delivered better.
Running one automation across both platforms at once
One connection covers both, because the Instagram professional account is linked to the page. Keep the offer and the keyword identical across the two, let each surface publish its own public reply, and log the results separately.
Identical keyword matters for a reason that only shows up later. If you use GUIDE on Instagram and TEMPLATE on Facebook for the same offer, you can no longer compare them, and you will spend months unable to say which surface produces buyers.
The public reply should differ slightly, though. On Instagram it has to mention the message requests folder, because a non follower will not find the message otherwise. On Facebook that sentence is noise, and noise in a short public reply is expensive.
Separate logs are the part people skip. A single combined number tells you the mechanic works. Two numbers tell you that Facebook delivered nearly everything it triggered while Instagram delivered half, which is an actionable difference. SocialFlow records each triggered comment with its platform, post and keyword, and that per surface split is the thing to insist on whatever tool you use.
Start with whichever platform your existing enquiries already come from, get one offer working there, then copy it across unchanged. Running both from day one with two different offers is how people end up with four numbers and no conclusion.